1. Technical Field
The present invention relates generally to radio frequency and microwave power measurements, and more particularly to a circuit providing automatic frequency determination with power measurements in which frequency data obtained can be used to correct the power measurements.
2. Background Art
When making RF and microwave power measurements the user typically must perform two or three operations, depending on the type of sensor: First, the user must zero the sensor; second, with some meters the user must calibrate the sensor; and third, the user makes a measurement. The user then must apply frequency correction (calibration factor) to the measurement. In older meters the user reads the calibration factor from a table or graph and then selects the calibration factor on the meter. Newer power meters allow the user to enter the frequency directly. The power meter firmware then selects the calibration factor for the user.
There has been at least one departure from this sequence. In the early to mid 1980's, the Hewlett-Packard Company created the HP8902A. The HP8902A combined in a single instrument a modulation analyzer, a power meter, and a very sensitive synchronous receiver/detector. This instrument was able to make frequency-corrected RF power measurements from +20 dBm to −120 dBm. However the meter still needed to be zeroed and calibrated by the user. The present inventors believe that this instrument was the first to measure the frequency automatically and then to automatically apply the power meter correction factors for the user. This was done by placing an RF switch, the essence of an HP8478A power sensor, and sufficient memory to hold the calibration factors for the sensor in a single unit.
This single unit was an HP11792A. The HP11792A connected directly to the HP8902A via two cables. The RF switch in the HP11792A provided two paths. One path routed the incoming signal to the power sensor through one cable. The second path routed the RF power to the modulation analyzer portion of the instrument.
When automatic measurements were made the switch routed the RF signal first to the modulation analyzer which measured the frequency. Then the switch in the HP11792A was set to route the RF signal directly to the HP8478A power sensor. Having measured the frequency the power reading could be automatically corrected.
Since the introduction of the HP8902A, it has become the norm for calibration factors to be uploaded automatically. And even though the HP8902A was very successful, automatic frequency-corrected power measurements have remained an uncommon feature. There may be a number of reasons for this: (1) The frequency determination has required a complex device (modulation receiver); (2) the resulting sensor package (HP11792A) was large and expensive; (3) in the case of the HP11792A, the switch was inclined to wear out over time; and (4) the need to constantly calibrate and zero the sensor minimized the advantage of automatic frequency determination.
In any event, the automatic application of calibration factors as a result of measuring frequency is not a novel idea. However, there have been ideas about how to measure the frequency more accurately. Most have often involved traditional approaches, such as frequency counters. These tend to have high accuracy and great resolution. Some include zero crossing detectors. However, these approaches have the following problems: (1) they tend to be slow; (2) they are very complex and expensive (usually as complex as the power meter itself); and (3) they are difficult to employ at low power and high frequency.
One important fact to understand when making frequency-corrected power measurements is that the frequency need not be measured with great precision—a good approximation is sufficient. For most purposes it is more than adequate if the frequency is known to within one percent (1%). Indeed, some approaches have recommended the use of frequency approximations.
At least one patent, U.S. Pat. No. 5,801,525, to Wiltron, seems to teach such an approach. The disclosed method has a few notable limitations: (1) Match is an important specification for any power sensor, and it is difficult to design filter that provides a broadband match to incoming signals (this would be problematic for power sensors); (2) while the filter is in the pass band it is difficult to reliably differentiate frequency based on level. Beyond the knee of the filter, sensitivity becomes a serious problem. These two limitations may make frequency determination too inaccurate in most cases, and simply impossible in some cases. This is because sensitivity varies with frequency (as previously mentioned), and as a result the useable power range is very limited. Next, as a practical matter greater frequency range (several decades) requires additional filters. Of course these will need to be switched in/out. This causes additional problems with match or reflection and complexity. In addition, if the switching is not conducted carefully, switching transients may be seen by the outside system. Further, as a practical matter it is necessary to employ one detector per filter. Each detector-filter pair must be calibrated as inferred by the patent. This results in an expensive and time consuming proposition since the filter-detector pair must be calibrated over power and temperature. And further still, the complexity of the firmware is greatly increased. In addition the time to make power measurements is increased. This is because it may be necessary to switch between all filters to determine the best signal level to use and then make the appropriate choice with respect to frequency.
Other approaches for determining frequency are common in literature related to the art. These include tuned notch filters, frequency discriminators and receivers. These approaches are either excessively complex, have very serious match problems, are unable to deal with a wide dynamic range, or suffers another fundamental limitation.
The foregoing patent and devices in the prior art reflect the current state of the art of which the present inventors are aware. Reference to, and discussion of, these patents is intended to aid in discharging Applicants' acknowledged duties of candor in disclosing information that may be relevant to the examination of claims to the present invention. However, it is respectfully submitted that none of the above-indicated patents disclose, teach, suggest, show, or otherwise render obvious, either singly or when considered in combination, the invention described and claimed herein.